Regional trends and the impact of various patient and hospital factors on outcomes and costs of hospitalization between academic and nonacademic centers after deep brain stimulation surgery for parkinson's disease: A United States Nationwide Inpatient Sample analysis from 2006 to 2010

Object. The aim of this study was to analyze the incidence of adverse outcomes, complications, inpatient mortal-ity, length of hospital stay, and the factors affecting them between academic and nonacademic centers after deep brain stimulation (DBS) surgery for Parkinson's disease (PD). The authors also analyzed the impact of various factors on the total hospitalization charges after this procedure. Methods. This is a retrospective cohort study using the Nationwide Inpatient Sample (NIS) from 2006 to 2010. Various patient and hospital variables were analyzed from the database. The adverse discharge disposition and the higher cost of hospitalization were taken as the dependent variables. Results. A total of 2244 patients who underwent surgical treatment for PD were identified from the database. The mean age was 64.22 ± 9.8 years and 68.7% (n = 1523) of the patients were male. The majority of the patients wasdischarged to home or self-care (87.9%, n = 1972). The majority of the procedures was performed at high-volume centers (64.8%, n = 1453), at academic institutions (85.33%, n = 1915), in urban areas (n = 2158, 96.16%), and at hospitals with a large bedsize (8.6%, n = 1907) in the West or South. Adverse discharge disposition was more likely in elderly patients (OR > 1, p = 0.011) with high comorbidity index (OR 1.508 [95% CI 1.148-1.98], p = 0.004) and those with complications (OR 3.155 [95% CI 1.202-8.279], p = 0.033). A hospital with a larger annual caseload was an independent predictor of adverse discharge disposition (OR 3.543 [95% CI 1.781-7.048], p < 0.001), whereas patients treated by physicians with high case volumes had significantly better outcomes (p = 0.006). The median total cost of hospitalization had increased by 6% from 2006 through 2010. Hospitals with a smaller case volume (OR 0.093, p < 0.001), private hospitals (OR 11.027, p < 0.001), nonteaching hospitals (OR 3.139, p = 0.003), and hospitals in the West compared with hospitals in Northeast and the Midwest (OR 1.885 [p = 0.033] and OR 2.897 [p = 0.031], respectively) were independent predictors of higher hospital cost. The mean length of hospital stay decreased from 2.03 days in 2006 to 1.55 days in 2010. There was no difference in the discharge disposition among academic versus nonacademic centers and rural versus urban hospitals (p > 0.05).Conclusions. Elderly female patients with nonprivate insurance and high comorbidity index who underwent surgery at low-volume centers performed by a surgeon with a low annual case volume and the occurrence of post-operative complications were correlated with an adverse discharge disposition. High-volume, government-owned academic centers in the Northeast were associated with a lower cost incurred to the hospitals. It can be recommended that the widespread availability of this procedure across small, academic centers in rural areas may not only provide easier access to the patients but also reduces the total cost of hospitalization.(http://thejns.org/doi/abs/10.3171/2013.8.FOCUS13295).